Introduction
Darwin developed the theory of sexual selection to explain evolution of traits, such as elaborated antlers or feathers, which by hampering survival, and thus apparently contradicted his theory of natural selection (Darwin 1859, 1871). Darwin’s solution highlighted the role of such costly traits in reproductive competition: antlers help males to win in combat over access to females, and colourful feathers make males sexually more attractive. However, he also appreciated that sexual competitiveness is likely to be associated with general health and vigour, and thus sexual selection may be partly aligned with natural selection (Darwin, 1859). In contemporary theory of sexual selection, this is framed in terms of condition-dependence of sexual trait expression, with condition being defined as a trait capturing genetic variation in an individuals’ ability to acquire and effectively process resources. This “genic capture” mechanism implies that individuals which are better-adapted and/or less burdened with mutations, should be more likely to reproduce and pass their genes to next generations (Andersson, 1986; Rowe & Houle, 1996).
The degree to which sexual selection is aligned with natural selection is likely to affect the risk of extinction. On the one hand, studies manipulating opportunity for sexual selection experimentally have demonstrated that sexual selection can help adaptation to a novel environment (Long et al., 2012; Plesnar-Bielak et al., 2012; Grieshop et al., 2016; Parrett & Knell, 2018), purge genetic load (Radwan, 2004; McGuigan et al., 2011; Almbro &Simmons, 2014) and in consequence, prevent extinction (Jarzębowska & Radwan, 2010; Plesnar-Bielak et al., 2012; Lumley et al., 2015; Godwin et al., 2020) (but see e.g. Arbuthnott & Rundle, 2012; Chenoweth et al., 2015, for counterexamples). On the other hand, elaborated sexual traits may become a burden to survival, especially during stress associated with environmental changes, which can eventually lead to extinction (Kokko & Brooks, 2003). While expression of condition-dependent sexually selected traits may be supressed under environmental stress, thus alleviating their costs to their bearers, genes underlying these traits may also have negative pleiotropic effects in female productivity and survival (Harano et al., 2010; Plesnar-Bielak et al., 2014; Łukasiewicz et al. 2020), which could contribute to increased risk of extinction (Kokko & Brooks, 2003).
Comparative work yielded conflicting results: some studies reported increased risk associated with elaboration of male sexually selected traits (Doherty et al., 2003; Morrow & Pitcher, 2003; Martins et al., 2018; Bro-Jørgensen, 2014), while others found no evidence for such relationship (Morrow & Fricke, 2004). These apparent inconsistencies can result from uncontrolled differences in the genetic diversity of populations of their exposure to environmental stressors. This possibility is highlighted by a recent study by Parrett et al. (2019), who found that generally positive effect of sexually selected beetle horns on survival is modulated by a degree of anthropogenic alteration of the habitat. Additional confounding factors may include differences in population sizes (Martinez-Ruiz & Knell, 2017) or in breadths of ecological niches, which can be modulated by the degree of sexual dimorphism (Bonduriansky, 2011; De Lisle & Rowe, 2015). Further progress could be achieved by using an experimental evolution approach, which have grossly contributed to our understanding of the contribution of sexual selection to population fitness (see Cally et al., 2019 for review). However, experimental evolution studies that measured extinction rate so far primarily manipulated mating systems or sex ratio (Plesnar-Bielak et al., 2012; Jarzębowska & Radwan, 2010; Parrett & Knell, 2018; Goodwin et al., 2020), rather than expression of exaggerated sexual traits. Thus, they cannot inform us on how evolution of such traits affects the risk of extinction.
Here, we experimentally assess the risk of extinction associated with the expression in males of a costly, sexually selected weapon under environmental change (gradual increase in ambient temperature by 20C per generation). Our model species was the male-dimorphic bulb mite Rhizoglypus robini , in which some males (but not females) express a sexually selected weapon: thickened third pair of legs. The weapon, expressed only in a fighter male morph is both significantly heritable and costly to produce (Radwan, 1995; Smallegange & Coulson, 2011). Part of the heritability could be due to higher load of deleterious mutations preventing males from expressing costly weapon. Indeed, comparison of inbreeding depression between inbred lines derived from fighters and scramblers indicated that the load of deleterious mutations is higher in the scramblers (Łukasiewicz et al., 2020). Thus, purifying selection should be stronger in populations in which the male weapon is present, possibly decreasing their risk of extinction. Furthermore, condition-dependence of the weapon expression (Radwan, 1995; Smallegange, 2011) should ensure that its direct costs to males can be plastically reduced under environmental challenge. However, females from genetic lines nearly fixed for fighter morph were shown to have lower fitness compared to lines nearly fixed for scrambler morph in two independent populations (Plesnar-Bielak et al., 2014; Łukasiewicz et al., 2020), indicating that genes associated with fighter phenotype can have negative pleiotropic effects on female fitness. This variant of gender load (Rice, 1992; Arnqvist & Tuda, 2010) may compromise population viability and make them more prone to extinction.
We used inbred lines nearly fixed for fighter or scrambler morph to establish outbred populations enriched for fighter or scrambler genes (F and S population respectively), each founded by the same number of genomes to control for initial genetic diversity. These populations were then subjected to an incremental increase in temperature (2°C per generation) to simulate a climate change comparable to that experienced by many species during current global warming (Parrett et al., 2018). We find that under such environmental change, populations enriched for fighter genes faced a significantly elevated risk of extinction. Mortality increased with the level of thermal stress at significantly faster rate among individuals in F populations, but males were not disproportionately more affected than females. Our results therefore imply that evolution of exaggerated sexual traits may have negative pleiotropic effect on fitness of both sexes under environmental challenge, thus increasing the risk of population extinction.